If the $selector is an node or a list of nodes all descendants that
match that node/node list are returned.
self::CONTEXT_DOCUMENT will use the document element as context. Otherwise the currently
nodes are used as contexts
self::FIND_MODE_FILTER will use the $selector only as filter and not execute it directly.
this is the like the jQuery specification - but a lot slower and needs more memory.
Additionally in this mode, it will find only element nodes. *
/**
* Discover a list of possible xpaths that are collection items
* in sorted order where the first element is the most likely.
* @return array[string]
*/
public function discoverScores(Nodes $nodes)
{
$nonContentNodes = $nodes->find('//*[not(text())]');
$maxSibs = (new Utils())->getMaxSibCount($nodes->getDocument(), $nonContentNodes);
arsort($maxSibs);
$ancestorCountGrouping = [];
foreach ($maxSibs as $path => $count) {
if ($count < 2) {
continue;
}
$ancestorCount = substr_count($path, '/');
if (!isset($ancestorCountGrouping[$ancestorCount])) {
$ancestorCountGrouping[$ancestorCount] = [$path => $count];
} else {
$ancestorCountGrouping[$ancestorCount][$path] = $count;
}
}
ksort($ancestorCountGrouping);
$result = [];
foreach ($ancestorCountGrouping as $_ => $collectionScores) {
foreach ($collectionScores as $path => $score) {
$result[$path] = $score;
}
}
return $result;
}
